Global population structure of Aspergillus terreus inferred by ISSR typing reveals geographical subclustering
© Neal et al; licensee BioMed Central Ltd. 2011
Received: 2 May 2011
Accepted: 16 September 2011
Published: 16 September 2011
Aspergillus terreus causes invasive aspergillosis (IA) in immunocompromised individuals and can be the leading cause of IA in certain medical centers. We examined a large isolate collection (n = 117) for the presence of cryptic A. terreus species and employed a genome scanning method, Inter-Simple Sequence Repeat (ISSR) PCR to determine A. terreus population structure.
Comparative sequence analyses of the calmodulin locus revealed the presence of the recently recognized species A. alabamensis (n = 4) in this collection. Maximum parsimony, Neighbor joining, and Bayesian clustering of the ISSR data from the 113 sequence-confirmed A. terreus isolates demonstrated that one clade was composed exclusively of isolates from Europe and another clade was enriched for isolates from the US.
This study provides evidence of a population structure linked to geographical origin in A. terreus.
Aspergillus species are believed to be cosmopolitan organisms, existing as unstructured global populations. Species belonging to this taxon, including A. fumigatus, A. terreus, A. flavus and others, cause invasive aspergillosis (IA) predominantly in severely immunocompromised individuals. The majority of studies with A. fumigatus have demonstrated no association between genotypes and geography. Several studies employing comparative sequence analysis of different loci, including protein coding, intergenic and microsatellite containing regions, arrived at the conclusion that there was no correlation between genotype and geographical origin among A. fumigatus isolates [1–3]. In contrast to these observations, one study demonstrated the presence of multiple, well-supported phylogenetic clusters amongst A. fumigatus isolates from a collection of isolates geographically dispersed across North America . The locus sequenced was a single gene encoding a putative cell surface protein, Afu3g08990 (CSP), in which polymorphisms consisted of insertions and deletions within a repeat region. The authors speculated that the presence of clusters may have been undetected previously due to the reliance on data from loci lacking sufficient polymorphisms.
Aspergillus terreus is the second or third most common etiological agent of IA and interestingly, appears to be the most common cause of infection in some medical centers, suggesting ecological specificity for this organism [5–7]. Previous efforts to determine population structure in A. terreus have been hampered by the lack of reliable methods for exploiting genetic variability to distinguish or group isolates. Balajee et al., employing a multi-gene sequencing approach to a large global collection of isolates previously identified as A. terreus, showed that no evidence of endemism existed but were able to define a genotypically distinct species, A. alabamensis . The use of multi-locus comparative sequence analysis to explore population structure in A. terreus supported the existence of a single globally distributed population . On the other hand, multiple studies using molecular fingerprinting methods, including RAPD, demonstrated high genotypic diversity among A. terreus isolates [9, 10], with no evidence of endemism [9, 11]. Thus, even as new species are defined within groups of isolates identified as A. terreus, support for the idea that A. terreus exists as a single, genotypically diverse, global population, lacking phylogeographic structure, continues [8–10].
A recent study investigating amphotericin B (AMB) susceptibility of a worldwide A. terreus collection found that isolates recovered from different parts of the world had different patterns of AMB susceptibility . At that time, no attempt was made to study the association between genotypic relatedness and antifungal susceptibility in this set of isolates. In the present investigation, this A. terreus isolate collection was genotyped employing the highly discriminatory genome-wide DNA fingerprinting method, Inter-Simple Sequence Repeat (ISSR) PCR  to (a) assess the use of this fingerprinting method for discriminatory genotyping of A. terreus; (b) evaluate the association between AMB susceptibility and genotype in this global collection of isolates; and (c) attempt to map geography onto genotypically related clusters of isolates. Results of this study revealed the possible global sub-structuring of genotypes and the presence of the recently described cryptic species A. alabamensis in Italy.
Fungal Strains and genomic DNA Isolation
A total of 117 clinical A. terreus isolates originating from France or Belgium (28 isolates), Italy (46 isolates), and the Eastern (22 isolates) and Western (21 isolates) United States were available for analyses from the previously performed study . All isolates were subcultured on Sabouraud Dextrose Agar (SDA) plates in preparation for genomic DNA isolation.
For genomic DNA extraction, fungal material was removed from plates and disrupted using an Omni mixer (Omni International, Warrenton, VA) in the presence of ATL buffer from the DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA) containing 1 mg/ml proteinase K (Sigma, St. Louis, MO). The disrupted material was incubated at 55°C for one hour with vortexing every 15 min. DNA was isolated using the DNeasy Blood and Tissue Kit (Qiagen, Valencia, CA) according to the manufacturer's protocol. Genomic DNA quality was checked with electrophoresis in a 1% agarose gel (Roche, Manheim, Germany) and quantity was measured with the nanodrop spectrophotometer at a wavelength of 260A (Thermo Fisher Scientific, Pittsburgh, PA).
Comparative Sequence Analysis of the calmodulin gene
Portions of the calmodulin locus (calM) were PCR amplified and sequenced as previously described . The resultant nucleotide sequences were edited with SeqMan Pro Ver 8.0.2 software (DNASTAR, Inc., Madison, WI). Edited sequences were aligned using Bioedit Sequence Alignment Editor Ver 188.8.131.52. Reference sequences were downloaded from GenBank and the software program GARLI [Genetic Algorithm for Rapid Likelihood Inference] was used to generate the maximum likelihood (ML) tree .
Development of ISSR Fingerprinting Method
ISSR primers designed for this study
Initially, ten ISSR primers were tested for their ability to generate reproducible, complex fingerprinting patterns on a panel of 40 A. terreus isolates randomly selected from the global isolate collection. For PCR amplification, 3-5 μl of genomic DNA was used as the template in a final reaction volume of 25 μl consisting of PCR buffer (10 mM Tris-HCl, 1.5 mM MgCl2, 50 mM KCl, pH 8.3); 0.2 mM each of dATP, dGTP, dCTP, and dTTP; 2 pmol of a single primer; and 1.3 U of Taq DNA polymerase (Roche Applied Science, Mannheim, Germany). Amplification was performed in a GeneAmp PCR system 9700 thermocycler (Applied Biosystems, Carlsbad, CA). Initial denaturation at 95°C for 5 min was followed by 36 cycles of 95°C for 30 s, 50°C for 45 s, and 72°C for 2 min. The last cycle was followed by a final extension at 72°C for 7 min. Fluorescently labeled PCR products were separated by capillary electrophoresis on an ABI 3130 DNA analyzer (Applied Biosystems, Carlsbad, CA). Briefly, 0.5 μl of a 1:10 dilution of PCR product was added to 0.25 μl GeneScanTM 1200 LIZ internal size standard and 9.25 μl Hi-Di formamide (Applied Biosystems, Carlsbad, CA). The 10 μl samples were denatured by heating to 95°C for 3 min., cooled and run on a 50 cm array in the POP-7 polymer matrix using the 1200LIZ run module.
Four of the ten primers tested produced complex, reproducible, banding patterns over multiple PCR reactions and a series of DNA concentrations, and these four ISSR primers were therefore selected for the analysis of the remaining sequence-confirmed A. terreus isolates.
Subtyping of the global A. terreus isolates
Fingerprints for all of the sequence-confirmed A. terreus isolates were generated using four ISSR primers that were selected after initial screening as described above. GeneMapper v4.0 (Applied Biosystems, Carlsbad, CA) was used to assign fragment sizes to the PCR products. Fragments identified using GeneMapper software were converted to binary data with a "0" representing the absence and a "1" representing the presence of an allele. The binary strings of data representing the fingerprint generated by each primer were concatenated in Excel (Microsoft Corporation, Redmond, WA) to form a single, continuous, binary string incorporating the results from all primers. Alleles that appeared in all or fewer than 10% of isolates were excluded from the analysis. Phylogenetic trees and Bayesian clusters were generated from identical binary data sets.
Phylogenetic Analysis of ISSR data
Neighbor-joining (NJ) trees were generated by PAUP [Phylogenetic Analysis Using Parsimony (and Other Methods)] . PHYLIP [Phylogeny Inference Package]  was used to produce the parsimony tree. Bayesian clustering was performed using the program STRUCTURE .
ISSR Fingerprinting of the Global A. terreus Isolates
On testing ten ISSR primers using a subset of forty A. terreus isolates, it was found that four primers were suitable for generating robust fingerprints for A. terreus: three trinucleotide repeat flanking primers and a single tetranuclotide repeat flanking primer (ISSR 7, 9, 10 and 13 respectively) (Table 1). These four ISSR primers were used to generate fingerprints for all of the sequence-confirmed A. terreus isolates. The A. alabamensis isolates were not fingerprinted. ISSR subtyping of 113 A. terreus revealed 111 unique genotypes with only two isolates, both from the same center in the Eastern United States, demonstrating identical fingerprinting patterns.
A significant relationship existed between geography and cluster membership (X2 = 48.2, d.f. = 6, p < 0.001), driven primarily by cluster 1 being composed only of isolates from Europe, as well as cluster 2 accounting for the majority [12 of 19] of the sequence-confirmed Eastern U.S. A. terreus isolates (Figure 2). The patterns of cluster membership in the two US populations were similar to each other and quite different from the pattern shared by the two European populations (Figure 2). There were nine isolates in which the majority contribution from any cluster was less than 0.66, suggesting that these isolates did not consistently fall into any one cluster. These isolates were excluded from any single cluster due to their high levels of inferred admixture.
Susceptibility testing to AMB
Mean MIC for Structure Defined Clusters
MEAN MIC (AMB) mg/L (→)
Extensive genotypic diversity has long been known in A. terreus, and recently a cryptic species, A alabamensis, was discovered among isolates originally identified as A. terreus . In the current study, we report the presence of four A. alabamensis isolates, identified by comparative sequence analysis of a previously characterized single locus gene (calM), from a collection of clinical isolates defined as A. terreus. Three A. alabamensis isolates were recovered from North America and one originated from Italy, making this the first reported A. alabamensis isolate recovered outside of North America. In contrast to a previous study that found that A. alabamensis had decreased in vitro susceptibility to AMB , all four A. alabamensis isolates recovered in this study had similar MIC patterns against AMB as compared to A. terreus (data not shown). None of the A. alabamensis isolates recovered in this study were colonizers (David Stevens, personal communication), a finding that was different from the study of Balajee et al. .
It has been postulated that unique A. terreus genotypes may occupy particular environmental niches associated with certain geographical areas. To test this hypothesis, Lass-Florl et al.  conducted a molecular epidemiological study using RAPD which explored the genotypes of clinical isolates recovered from two medical centers that more frequently reported A. terreus infections. Results of this study reported a great diversity of genotypes among isolates from both centers and revealed no evidence of endemicity among the isolates at either center. Another study investigating in vitro activity of AMB against a large global collection of clinical isolates suggested that isolates from different parts of the world could have differences in AMB susceptibility . Tortorano et al.  found that of the four geographic locations where isolates originated, the average MIC of the isolates from the Eastern United States were statistically different from those of the isolates from the other three geographical regions namely, Italy, France-Belgium, and the Western United States, suggesting a possible association between geography and MIC. We postulated that an association between geography and genotype may also exist in this population and that ISSR fingerprinting may be a superior way to assess population structure in these isolates.
Results of our study demonstrated high genotypic diversity within these isolates with only two isolates displaying identical fingerprinting patterns. In spite of this high genotypic diversity, sufficient common markers existed between isolates to group them into distinct clades supported by multiple phylogenetic methods. Specifically, Bayesian clustering in the program STRUCTURE revealed 3 distinct clusters of isolates that were in agreement with the clades inferred by NJ. Cluster 2 (Figure 2 and 3) generated by STRUCTURE shares the isolates in clade 2 of the NJ tree which had the highest bootstrap support of any clade. This suggests that these isolates share alleles that are less enriched in isolates from the other two clades, and thus may be the most ancient group. Isolates in cluster 1 were restricted to Europe, while isolates in cluster 2 were most commonly recovered from the U.S., and cluster 3 included isolates recovered globally. There were nine isolates with high levels of inferred admixture that did not belong to any single cluster. It is tempting to speculate that human activities may have facilitated the global distribution of cluster 3 and the admixture between populations.
Clustering of isolates from the same sampling area suggests a link between genetic similarity and geographic origin in a population of organisms previously believed to lack endemism. Additional isolates from both clinical and environmental sources obtained from diverse geographical regions will need to be rigorously examined to verify the endemism suggested by our study. An expanded population structure analysis including isolates with more complete epidemiological data could lend predictive power about antifungal susceptibility to future studies. In contrast to the above finding, the relationship between population structure and AMB susceptibility was small. This could be attributable to the sample size being too small or to the lack of an association between in vitro antifungal susceptibilities and geographical origin.
Multiple studies have demonstrated that A. terreus is the predominant etiological agent of IA in certain medical centers around the world including those in Houston, Texas, and Innsbruck, Austria [5, 9, 18]. Molecular examination of isolates from these centers showed no endemism and the authors concluded that other factors including levels of immunosuppression and previous antifungal use in the host, could, in part, be responsible for the prevalence of A. terreus in these medical centers. We have demonstrated in this study, using a discriminatory molecular method, a different set of globally derived isolates and rigorous phylogenetic analysis of the resulting data, that A. terreus may exhibit endemism. It remains to be seen if isolates recovered from other parts of the world, importantly from centers in Texas and Austria, cluster according to geographical location. Such molecular characterization of medically important fungi could be pivotal in understanding the ecology, acquisition and transmission of these organisms.
The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the CDC.
The authors would like to thank Oliver Clay and Sujatha Seenu for providing assistance with PERL and R scripts. Carolyn Neal was supported by the Emerging Infectious Diseases Fellowship sponsored by the Association of Public Health Laboratories and the Centers for Disease Control and Prevention.
- Bain JM, Tavanti A, Davidson AD, Jacobsen MD, Shaw D, Gow NA, Odds FC: Multilocus sequence typing of the pathogenic fungus Aspergillus fumigatus. J Clin Microbiol. 2007, 45 (5): 1469-1477. 10.1128/JCM.00064-07.PubMedPubMed CentralView ArticleGoogle Scholar
- Rydholm C, Szakacs G, Lutzoni F: Low genetic variation and no detectable population structure in aspergillus fumigatus compared to closely related Neosartorya species. Eukaryot Cell. 2006, 5 (4): 650-657. 10.1128/EC.5.4.650-657.2006.PubMedPubMed CentralView ArticleGoogle Scholar
- Pringle A, Baker DM, Platt JL, Wares JP, Latge JP, Taylor JW: Cryptic speciation in the cosmopolitan and clonal human pathogenic fungus Aspergillus fumigatus. Evolution Int J Org Evolution. 2005, 59 (9): 1886-1899.View ArticleGoogle Scholar
- Balajee SA, Tay ST, Lasker BA, Hurst SF, Rooney AP: Characterization of a novel gene for strain typing reveals substructuring of Aspergillus fumigatus across North America. Eukaryot Cell. 2007, 6 (8): 1392-1399. 10.1128/EC.00164-07.PubMedPubMed CentralView ArticleGoogle Scholar
- Lass-Florl C, Griff K, Mayr A, Petzer A, Gastl G, Bonatti H, Freund M, Kropshofer G, Dierich MP, Nachbaur D: Epidemiology and outcome of infections due to Aspergillus terreus: 10-year single centre experience. Br J Haematol. 2005, 131 (2): 201-207. 10.1111/j.1365-2141.2005.05763.x.PubMedView ArticleGoogle Scholar
- Lass-Florl C, Rath P, Niederwieser D, Kofler G, Wurzner R, Krezy A, Dierich MP: Aspergillus terreus infections in haematological malignancies: molecular epidemiology suggests association with in-hospital plants. J Hosp Infect. 2000, 46 (1): 31-35. 10.1053/jhin.2000.0799.PubMedView ArticleGoogle Scholar
- Baddley JW, Pappas PG, Smith AC, Moser SA: Epidemiology of Aspergillus terreus at a university hospital. J Clin Microbiol. 2003, 41 (12): 5525-5529. 10.1128/JCM.41.12.5525-5529.2003.PubMedPubMed CentralView ArticleGoogle Scholar
- Balajee SA, Baddley JW, Peterson SW, Nickle D, Varga J, Boey A, Lass-Florl C, Frisvad JC, Samson RA: Aspergillus alabamensis, a new clinically relevant species in the section Terrei. Eukaryot Cell. 2009, 8 (5): 713-722. 10.1128/EC.00272-08.PubMedPubMed CentralView ArticleGoogle Scholar
- Lass-Florl C, Grif K, Kontoyiannis DP: Molecular typing of Aspergillus terreus isolates collected in Houston, Texas, and Innsbruck, Austria: evidence of great genetic diversity. J Clin Microbiol. 2007, 45 (8): 2686-2690. 10.1128/JCM.00917-07.PubMedPubMed CentralView ArticleGoogle Scholar
- Blum G, Perkhofer S, Grif K, Mayr A, Kropshofer G, Nachbaur D, Kafka-Ritsch R, Dierich MP, Lass-Florl C: A 1-year Aspergillus terreus surveillance study at the University Hospital of Innsbruck: molecular typing of environmental and clinical isolates. Clin Microbiol Infect. 2008, 14 (12): 1146-1151. 10.1111/j.1469-0691.2008.02099.x.PubMedView ArticleGoogle Scholar
- Symoens F, Bouchara JP, Heinemann S, Nolard N: Molecular typing of Aspergillus terreus isolates by random amplification of polymorphic DNA. J Hosp Infect. 2000, 44 (4): 273-280. 10.1053/jhin.1999.0707.PubMedView ArticleGoogle Scholar
- Tortorano AM, Prigitano A, Dho G, Biraghi E, Stevens DA, Ghannoum M, Nolard N, Viviani MA: In vitro activity of amphotericin B against Aspergillus terreus isolates from different countries and regions. J Chemother. 2008, 20 (6): 756-757.PubMedView ArticleGoogle Scholar
- Cano J, Rezusta A, Sole M, Gil J, Rubio MC, Revillo MJ, Guarro J: Inter-single-sequence-repeat-PCR typing as a new tool for identification of Microsporum canis strains. J Dermatol Sci. 2005, 39 (1): 17-21. 10.1016/j.jdermsci.2005.01.016.PubMedView ArticleGoogle Scholar
- Zwickl DJ: GARLI Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. 2006, Austin: The University of Texas at AustinGoogle Scholar
- Swofford DL: PAUP* 4.0: phylogenetic analysis using parsimony (*and other methods). 1999, Sunderland, Massachusetts: Sinauer Associates, Inc., 4.0b2aGoogle Scholar
- Felsenstein J: PHYLIP (Phylogeny Inference Package) version 3.68. 1993, Department of Genome Sciences, University of Washington, SeattleGoogle Scholar
- Pritchard J, Stephens M, Donnelly P: Structure. 2000, Department of Statistics, University of Oxford, Oxford, United Kingdom, v. 2.3.3Google Scholar
- Hachem RY, Kontoyiannis DP, Boktour MR, Afif C, Cooksley C, Bodey GP, Chatzinikolaou I, Perego C, Kantarjian HM, Raad II: Aspergillus terreus: an emerging amphotericin B-resistant opportunistic mold in patients with hematologic malignancies. Cancer. 2004, 101 (7): 1594-1600. 10.1002/cncr.20554.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.